THE EFFECTIVENESS OF STRATIFIED CONSTRUCTED WEEK SAMPLING FOR CONTENT ANALYSIS OF ELECTRONIC NEWS SOURCE ARCHIVES: AP NEWSWIRE, BUSINESS WIRE, AND PR NEWSWIRE

This study extends the content analysis sampling literature through exploration of sample size requirements for electronic news source archives. Significantly larger samples are required to achieve representativeness for AP Newswire, Business Wire, and PR Newswire than for more traditional sources of news content. When sampling press releases, constructing Weeks on a quarterly basis provides more representative samples than constructing weeks on a full-year basis for some categories, particularly those tied to a fixed business cycle. The results support the idea that as information passes through more media gatekeepers, who limit and standardize content, the sample sizes required for content analysis diminish.